The question here is, why is the sign wrong on power and energy?
In my new installation (7 Aug 26), it would be good if my water heater generated electric energy, but I am too addicted to reality to think that this is happening, despite living in California (read split phase).
The two 200A CTs are in the same direction with the wire side going towards the breakers and away from the grid. All other CTs are with the K side nearer the breaker and the L side at the load. Double and triple checked.
I put the water heater CT on the same wire as the oven ct. On the Setup–>Inputs page I see the P3 Oven with 1690 and the P6 Water Heater with -1355. I swap the CTs on the emonTx6 and the result is ~the same. That is, the problem seems to be with P6 of the emonTx6.
emonTx6-P2 Cooktop, and emonTx6-P4 Dryer, see the same behaviour,
Why is your emonVS shown connected to Leg 1 twice, on circuits 5 & 8? Surely, the “L2” input of the emonVS should be measuring Leg 2 of your supply? Or is this a typo, and you forgot to change one of those lines on the spreadsheet?
See Use in North America — OpenEnergyMonitor 0.0.1 documentation. Note the emonVs L2 input (V2) is connected to L2, not L1 as your spreadsheet says.
This points to an error with how the pairing of each current channel to link it to the correct voltage channel is set up for split-phase systems. Can you check this? The emonVS gets its operating power from L1, so if you pull the fuse or trip the L2 breaker feeding it, all the powers for circuits on L2, and only those, should go to zero. You have no 240 V circuits?
The pairing is done in the emonTx6 itself. You can check this by going to emonCMS Admin → Serial Config and connect to the emonTx6. Confusingly, this page seems to imply, with the choices available for V Chan 1 and V Chan 2, that every load is connected Line – Line (in your case, 240 V), and doesn’t make it clear what you have to do with a load connected Line – Neutral. In fact, you must put the same number in both boxes - 1 for a load connected to Leg 1, and 2 for a load connected to Leg 2. Having experimented, I think what you have here might be wrong.
If not, check for the silly mistake - make sure you have not slipped a minus sign into the calibration constants somewhere.
I don’t know what this means and can’t check it, because all I have for this item is the data sheet, and this gives no phasing information. But if the two incoming powers are positive when you are consuming, they are correct.
Thanks for the information @Robert.Wall ! It was a long hot day for me and the little gray cells got fried, not to mention banging my head a-good-one on the panel door … twice!
The information about the pairing is what I was seeking. I’ll let you know how it goes.
200A CTs My panel is not CT friendly for monitoring the grid. I put them in they way the fit and apparently, the wrong direction. I’ll sort this out once I get the pairing sorted.
Notes an North American panels
(bottom line: the emonVS is correctly connected)
North American panels are complex because of the split phase. Here I have highlighted how the phases connect to the breakers. The phase connections go down the center and are interleaved. The emonVS is connected to the outside two-pole (one pole trips, the other also trips) of a quad breaker (the EV connects to the inner two-pole) One leg of the emonVS is connected to L2 at point 7A and the other to L1 at 9B.
There are two ways to fix this without touching the c.t’s: swap the wires at the plug end (it’s quite safe to do this without removing the power, just unplug them, and it won’t hurt if the wires are shorted while you do it), or if there is no connection to L3 of your emonVs (and this was never intended in the original design ), put the number 3 for V Chan 1 for both incomers, leaving V Chan 2 at 1 for the one on Leg 1 and 2 for the one on Leg 2.
That’s not how I understood it, and not how I read your table. Maybe a proper single line diagram would have been better?
I have to say I’ve never come across a linked breaker quite like that one. Yes, for 3-phase operation we have 4-pole, or rather 3-pole & neutral, with all 4 poles linked but with overcurrent only on the 3 line connections.
Yes please. My 3-phase emonVs has the wrong plug (6P6C) for my emonPi3/emonTx6 board (socket is 8P8C), so I can’t rig up a split supply to test it. I did so to test the emonTx3 and emonLibDB, and it was all correct - but the pairing was done in the software directly and not via emonCMS in those days.
I’m a bit confused. There are has 12 ct’s The Hardware is emonPi3. I don’t know how to display the emonTx6. How is this done?
I made the effort to flip the L1 and L2 CTs to avoid confusion.
I try to make changes via the Serial Config: Admin–>Serial Config–>Stop EmonHub–>set to ttyAMA0–>Start I make a change, eg CT 1/V Chan 1 set to 3, save changes, stop (serial), Start EmonHub. But when I go back to Serial Config, the is no change. What did I miss?
This doesn’t show what I originally saw when I started up so I downloaded the emonPi3 image, copied it to a 32GB SD and it looked like it started OK, but I couldn’t connect to the WiFi because the password was wrong. I looked at the page emonSD Download → Credentials for passwords, but none worked. It seems that the password must be 8 characters. Is the password what is shown briefly when the emonPi3 first boots?
You need to connect it to the Pi via USB-A at the Pi end to USB-C at the emonTx6 end, and change the serial port emonCMS uses to suit - it should be ttyS0.
What did you miss? I don’t know. It worked for me twice, once setting the two c.t’s differently, and again setting them back again.
Is it your browser cache, by any chance? All I did differently to you each time was to go back to the Input page to check that emonHub had indeed started.
A new emonSD image does not have a Wi-Fi password. You can either create a file on the SD card immediately after you flashed it (and if needs be, expanded some partitions – to create the file see Setup WiFi on a Pi Manually using wpa_supplicant.conf - Raspberry Pi Spy and start at Step 2), or you need to set the emonPi as a Wi-Fi Access Point, which you can do with the front panel button. emonSD Download — OpenEnergyMonitor 0.0.1 documentation
The Wi-Fi password is the password which goes with the SSID of your Wi-Fi router, the same as you use when you configure your phone or laptop to access your router.
EmonCMS does not have a user account or password set when it first starts. You need to create both (“Register”) the first time you try to access emonCMS. Thereafter, you “Login” to emonCMS using this username and password. (By default, only a single user is allowed.)
I don’t think that is true, I have set longer passwords than this. I have not tested to see if 8 is the minimum size.
I figured out the password to WiFi problem. The last line of the “failed to connect” says to try a different method (click here). I clicked, another dialog box appeared and entered the password from the aforementioned page.
I re-imaged the SD card. The Serial Config page looks like it did before I re-imaged it. Am I right to assume that the data is stored in NVRAM?
I am able to change parameters and they seem to stay fixed.
When I change CT1 to V Chan 1 to 1 and V Chan 2 to 1, the power goes to 0W. The same results for CT2 setting to channel 2. Wait! CT1 power has gone to 1905! I measured the current at CT1 and CT2 and it is about 18A which is what would be expected, ~120*18=2160.
I tried 1 & 3 and 2 & 3 which also yielded 0W. What if the V3 in the emonVS were connected to the neutral? Would that achieve the desired results for this configuration?
When I click the Active box on V1 and V2, they stay clicked but the console reports No changes to save. A subsequent start-stop cycle unchecks the boxes.
The emonTx6 configuration has disappeared! (as reported previously).
I appreciate your help, Robert!!!
On a logistical note, I can only spend a few more hours on this project. I’ll be back late September.
Only briefly, to minimise the number of writes to the SD card. Eventually (the timescale is minutes), all the data you record in a Feed is stored in the Data partition (the large one) on the SD Card. If you write a downloaded image to a used SD card, all the data – absolutely everything – that was on the card is lost. If you need to preserve it, you must backup the data to another computer, and restore it to the new image.
I’m not sure I understand what emonCMS is doing in the background. The inherent assumption, which reflects the physical connections inside the emonVs, is the voltage is measured Line – Neutral. If a load is connected Line – Line, you need the voltage at both ends so you can subtract one from the other to get the voltage across the load. If @awjlogan has copied emonLibCM & emonLibDB in how it calculates power in his software, then needs only one voltage – of the Line the c.t. is drawing current from – to calculate the power, and so there should be nothing entered in the second voltage connection column. But you don’t have that option, only 1, 2 or 3 is offered, so it looks as if @TrystanLea had done a fiddle in emonCMS to say, in effect “If you say both ends of the load are connected to the same line, I’ll assume you mean Line – Neutral and instruct the measurement end accordingly”.
What I’m telling you to do is tell emonCMS the load is connected between L1 and L3, so it should use V1 & V3, and V3 always has zero volts on it (like the neutral) because it is not connected (but you could connect L3 to neutral and it should be exactly the same) and when you put 1 and 3 for the voltages, the maths will calculate the voltage of L1 with respect to L3. If you reverse that, it will calculate the voltage of L3 with respect to L1, which is the opposite sign, which will reverse the sign of the power! Which is what you want.
If you can wait, your problem has triggered a discussion and we’re looking at tidying this up to make it clearer and hopefully add a tickbox to reverse the power, so as to correct for a wrongly installed c.t.
It will, if you have unplugged the emonTx6 from the emonPi3. The emonPi3 needs to read the configuration data stored inside the emonTx6 to show it in emonCMS, then write it back when you press Save. This can’t be done over the air, it has to go by USB.
I connected a USB A-C cable from the emonPi3 USB-A to the USB-C connector of the emonTx6. When I ssh to the emonPi3, I can note that /dev/ttyACM0 now exists. When I go to Admin–>Serial Config (stop emonhub), ttyACM0 is not in the drop down box.
I thought Gregg was referring to the emonPi3 Serial Config as he doesn’t appear to be getting as far as the emonTx6, and because it immediately follows him saying he re-imaged the card, I took it to mean the Feeds data.
Why do you want to do this?
Yes, it is the connection between the Raspberry Pi and the measuring front end - the emonPi3/emonTx6 board inside your emonPi3 case.
The connection you want to see is the emonTx6 configuration is the one I told you above, the USB-A to USB-C lead connecting your emonPi3 to your emonTx6, that one is ttyS0. Select this, it is the second line in the drop-down box for the Serial connection on the Serial Config Tool page, and start. You should see a similar picture to the emonPi config but it will have the emonTx6 settings.
If you want to configure the emonTx6 directly from your computer, then you again must use the USB-C lead plugged in to the emonTx6, and then connect to a USB port on the computer and run a serial terminal tool on the computer. The details are here Configuration — OpenEnergyMonitor 0.0.1 documentation
If you do this, all you are doing is sending exactly the same commands to the emonTx6 as emonCMS sends.
@OldSurferDude
If you’re going to be away for 6 weeks, it’s probably best to wait and see if the emonCMS tool is revised and updated.
Just to be clear here, ACM0 is (likely) to be the wired USB connection (“Tx6”), S0 is the internal UART on the GPIO pins (“Pi3”). I say likely as the OS is free to rename, so just be careful here.
If you are purely using it in a box with a Raspberry Pi, I would avoid complication by using the internal UART on the GPIO pins. This is either through emonCMS, or through the terminal (ssh as described above). If going through the terminal, stop emonCMS first so there’s no contention for the UART.
The commands being sent are equivalent and can be interleaved, but let’s try to reduce sources for confusion by going through a single interface
That appears to be the opposite way round to what I see. Given that I can’t connect an emonTx6 (because I have only one and it’s an emonPi3 on a RPi 2B), I see the measurement front end when I connect to ttyAMA0 and I get no response from ttyS0.
[@awjlogan Is there a software version that I could load onto the first prototype emonPi3 board you sent (v0.1 (63dae4f)) without the RFM69CW, expansion header or LCD) that would allow me to replicate what Gregg is trying to do? ]
Yes, that’s the OS renaming them. Notice I mentioned ACM not AMA - annoyingly similar notation.
Unfortunately it would be woefully out of data and not functional. Let me see if I can make some changes to the current one - ideally it might just be changing some pin assignments…
), put the number 
